The FDA's recently established guidelines for the approval of prescription botanical drugs focuses on botanical drug safety, efficacy and product consistency. While safety and efficacy can be addressed in well designed clinical trials, defining metrics to measure product consistency of the complex collection of phytochemicals inherent to botanical extracts remains a major unsolved problem. Chemical characterization utilizing marker compounds, is today the standard for assessing quality control. However, this method is inadequate, as it cannot detect all chemical components nor address the issue of biological relevance or importance of any of the phytochemical components. Recent technology advances in transcription profiling and comprehensive gene microarrays provide a feasible pathway to develop a holistic, bioresponse profile that focuses not on the chemical components, but on the integrated biological response of a human cell to the entire composition of chemical components. The long term goal of this proposal is to develop a standardized FDA approved process for developing unique, diagnostic gene sets to characterize the biological response of botanical drugs as a metric to assess drug product reproducibility. Towards this goal we propose, the following two specific aims: 1. Creation of a pilot gene expression profile database for ten common western and Chinese botanicals 2. Utilize these gene expression profiles to build classification models to illustrate the utility in distinguishing different botanicals and to define a preliminary set of candidate genes for further development of a Quality Control metric for a PhytoCeutica botanical drug candidate PHY906 Successful completion of these goals will demonstrate the overall feasibility of utilizing gene expression patterns as a unique and sensitive bioresponse metric for botanical drug quality control. Phase II, will then focus on expanding and validating this technology for PHY906 and developing a proprietary gene chip as an FDA accepted method for a bioequivalent quality control measure.